Table 3. Beta coefficients (SE) for random intercept models assessing the association between rationing and sachet use among Accra women.
Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | ||||||
Characteristic | Empty (no covariates) | Full model of EA- and individual-level factors | Full model with level-2 interactions | Full model with level-2 and cross-level interactions | Full model with level-1 interactions | |||||
Intercept | −3.077*** | (0.149) | −5.133*** | (0.825) | −5.830*** | (0.942) | −6.061*** | (1.138) | −6.678*** | (1.807) |
Interactions: Level-2 (EA) | ||||||||||
Rationing×Vegetation | −1.420 | (0.875) | −1.648∼ | (0.889) | ||||||
Rationing×Upper middle class SES | −0.404∼ | (0.241) | −0.389 | (0.243) | ||||||
Rationing×Lower middle class SES | −0.072 | (0.230) | −0.047 | (0.238) | ||||||
Rationing×Lower class SES | −0.176 | (0.230) | −0.109 | (0.236) | ||||||
Interactions: Cross-level | ||||||||||
Rationing×OH Very good | −0.106 | (0.216) | ||||||||
Rationing×OH Good | −0.150 | (0.205) | ||||||||
Rationing×OH Fair or poor | −0.151 | (0.217) | ||||||||
Rationing×wealth score | 0.084∼ | (0.050) | ||||||||
Upper middle class SES×wealth score | 0.156 | (0.218) | ||||||||
Lower middle class SES×wealth score | −0.405 | (0.309) | ||||||||
Lower class SES×wealth score | 0.627** | (0.231) | ||||||||
Interactions: Level-1 (Individual) | ||||||||||
Wealth score×OH Very good | 0.213 | (0.634) | ||||||||
Wealth score×OH Good | 0.612 | (0.606) | ||||||||
Wealth score×OH Fair or poor | 0.896 | (0.628) | ||||||||
Age×OH Very good | −0.035∼ | (0.019) | ||||||||
Age×OH Good | −0.007 | (0.007) | ||||||||
Age×OH Fair or poor | −0.029** | (0.011) | ||||||||
Level-1 (Individual) variables | ||||||||||
Self-reported overall health | ||||||||||
Excellent† | ||||||||||
Very good | 1.264* | (0.502) | 1.259* | (0.502) | 1.593∼ | (0.822) | 2.955∼ | (1.745) | ||
Good | 1.853*** | (0.483) | 1.855*** | (0.484) | 2.312** | (0.796) | 2.610 | (1.646) | ||
Fair or poor | 2.151*** | (0.516) | 2.139*** | (0.517) | 2.633** | (0.833) | 3.988* | (1.711) | ||
Solid waste disposal | ||||||||||
Burnt, buried, dumped elsewhere, other† | ||||||||||
Public dump | −0.355 | (0.303) | −0.326 | (0.304) | −0.391 | (0.305) | −0.428 | (0.307) | ||
Collection service | 0.104 | (0.323) | 0.109 | (0.325) | 0.078 | (0.327) | 0.038 | (0.327) | ||
Liquid waste disposal | ||||||||||
Sewage system† | ||||||||||
Thrown in street, gutter, compound, other | 0.592∼ | (0.333) | 0.578∼ | (0.334) | 0.497 | (0.344) | 0.484 | (0.354) | ||
Major ethnic group | ||||||||||
Other | ||||||||||
Ewe | 0.024 | (0.300) | 0.015 | (0.310) | 0.004 | (0.304) | −0.019 | (0.309) | ||
Ga | −0.255 | (0.290) | −0.275 | (0.291) | −0.312 | (0.293) | −0.292 | (0.296) | ||
Akan | 0.012 | (0.269) | −0.007 | (0.270) | −0.011 | (0.273) | 0.004 | (0.275) | ||
Age (years) | −0.016** | (0.005) | −0.016** | (0.005) | −0.015** | (0.006) | – | (–) | ||
Wealth score | 0.380** | (0.125) | 0.392** | (0.126) | – | (–) | – | (–) | ||
Level-2 (EA) variables | ||||||||||
SES quartile | ||||||||||
Upper class† | ||||||||||
Upper middle class | −0.057 | (0.449) | 0.827 | (0.645) | 0.718 | (0.659) | 0.720 | (0.664) | ||
Lower middle class | 0.523 | (0.482) | 0.674 | (0.695) | 0.431 | (0.712) | 0.381 | (0.717) | ||
Lower class | 1.354** | (0.489) | 1.776* | (0.712) | 1.703* | (0.724) | 1.704* | (0.728) | ||
% Vegetated land cover | 2.069 | (1.290) | 4.881* | (1.964) | 5.452** | (1.988) | 5.464** | (2.010) | ||
Days of rationing | 0.087 | (0.069) | 0.395∼ | (0.233) | 0.514∼ | (0.311) | 0.531∼ | (0.312) | ||
Random Effects | ||||||||||
Level-2 variance (EA) | 2.040 | (0.368) | 1.691 | (0.349) | 1.636 | (0.354) | 1.668 | (0.361) | 1.711 | (0.368) |
Δ (%) in level-2 variance | – | −17.1% | −19.8% | −18.2% | −16.1% | |||||
−2 res log pseudo-likelihood | 15,945.99 | 15,791.81 | 15,905.96 | 16,083.04 | 16,285.13 | |||||
Generalized chi-square | 1,427.79 | 1,429.90 | 1,483.29 | 1,504.83 | 1,527.98 |
Reference Category;
∼ p<0.10;
p<0.05;
p<0.01;
p<0.001.